Prior research (with predominantly younger samples) established that patients with severe mental illnesses such as schizophrenia are at risk for impaired decision making capacity, yet even within diagnostic categories substantial variability exists among patients in the levels of such capacity. While there has been little research on decision making capacity (and it enhancement) in older psychiatric patients, older patients may have a more complex array of factors to consider in evaluating the risk-benefit ratio of participating in a particular research protocol, yet may also have a partially diminished capacity to understand, appreciate and reason about these factors. Also, while informed consent is ideally viewed as a ongoing process throughout a patient-participant's enrollment in a study, there has been little research on the degree to which decision making capacity may fluctuate over time. The proposed study is designed to help fill these gaps. Subjects will include 152 patients age 50 years or older with schizophrenia (n=76) or mood disorders with psychotic features (n=76) who are considering participation in a randomized longitudinal comparison of three atypical antipsychotic medications. All subjects will be evaluated at baseline with standardized measures of decision making capacity, severity of psychiatric symptoms, insight/awareness of illness, and specific cognitive-abilities. To examine the ability of patients' to profit from multiple learning trials, decision making capacity will be assessed three times in succession at each visit. Between each of the these three within session decision-making capacity assessments, any information that patients had difficulty with on the preceding trial will be represented. The decision making capacity measure will be re- administered one-week later to establish test-retest reliability, and all of assessments will be repeated at 3-month, 1-year, and 2-year followup evaluations (except that the cognitive battery will be a shorter version of the baseline battery, targeting those cognitive abilities which might be expected to change). We hypothesize that patients' levels of capacity within four specific dimensions of decision making (understanding, appreciation, reasoning, and evidence of a choice) will be differentially associated with specific cognitive skills, psychiatric symptoms, and levels of insight. We also expect that patients' decision making capacity will be significantly improve within sessions by multiple presentations of the material. We further hypothesize that fluctuations in capacity from baseline to the 3-month, 1-year, and 2-year followups will be accompanied by parallel changes in insight deficits and severity of psychiatric symptoms.